<feed xmlns='http://www.w3.org/2005/Atom'>
<title>linux-toradex.git/arch/powerpc/include/asm/processor.h, branch v3.10.58</title>
<subtitle>Linux kernel for Apalis and Colibri modules</subtitle>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/'/>
<entry>
<title>powerpc/tm: Fix context switching TAR, PPR and DSCR SPRs</title>
<updated>2013-08-15T05:59:05+00:00</updated>
<author>
<name>Michael Neuling</name>
<email>mikey@neuling.org</email>
</author>
<published>2013-08-09T07:29:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=ab9530578e03d7fc00f41f4c406d213d81579de0'/>
<id>ab9530578e03d7fc00f41f4c406d213d81579de0</id>
<content type='text'>
commit 28e61cc466d8daace4b0f04ba2b83e0bd68f5832 upstream.

If a transaction is rolled back, the Target Address Register (TAR), Processor
Priority Register (PPR) and Data Stream Control Register (DSCR) should be
restored to the checkpointed values before the transaction began.  Any changes
to these SPRs inside the transaction should not be visible in the abort
handler.

Currently Linux doesn't save or restore the checkpointed TAR, PPR or DSCR.  If
we preempt a processes inside a transaction which has modified any of these, on
process restore, that same transaction may be aborted we but we won't see the
checkpointed versions of these SPRs.

This adds checkpointed versions of these SPRs to the thread_struct and adds the
save/restore of these three SPRs to the treclaim/trechkpt code.

Without this if any of these SPRs are modified during a transaction, users may
incorrectly see a speculated SPR value even if the transaction is aborted.

Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;

</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
commit 28e61cc466d8daace4b0f04ba2b83e0bd68f5832 upstream.

If a transaction is rolled back, the Target Address Register (TAR), Processor
Priority Register (PPR) and Data Stream Control Register (DSCR) should be
restored to the checkpointed values before the transaction began.  Any changes
to these SPRs inside the transaction should not be visible in the abort
handler.

Currently Linux doesn't save or restore the checkpointed TAR, PPR or DSCR.  If
we preempt a processes inside a transaction which has modified any of these, on
process restore, that same transaction may be aborted we but we won't see the
checkpointed versions of these SPRs.

This adds checkpointed versions of these SPRs to the thread_struct and adds the
save/restore of these three SPRs to the treclaim/trechkpt code.

Without this if any of these SPRs are modified during a transaction, users may
incorrectly see a speculated SPR value even if the transaction is aborted.

Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;

</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc/tm: Fix userspace stack corruption on signal delivery for active transactions</title>
<updated>2013-05-31T22:29:23+00:00</updated>
<author>
<name>Michael Neuling</name>
<email>mikey@neuling.org</email>
</author>
<published>2013-05-26T18:09:41+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=2b3f8e87cf99a33fb6faf5026d7147748bbd77b6'/>
<id>2b3f8e87cf99a33fb6faf5026d7147748bbd77b6</id>
<content type='text'>
When in an active transaction that takes a signal, we need to be careful with
the stack.  It's possible that the stack has moved back up after the tbegin.
The obvious case here is when the tbegin is called inside a function that
returns before a tend.  In this case, the stack is part of the checkpointed
transactional memory state.  If we write over this non transactionally or in
suspend, we are in trouble because if we get a tm abort, the program counter
and stack pointer will be back at the tbegin but our in memory stack won't be
valid anymore.

To avoid this, when taking a signal in an active transaction, we need to use
the stack pointer from the checkpointed state, rather than the speculated
state.  This ensures that the signal context (written tm suspended) will be
written below the stack required for the rollback.  The transaction is aborted
becuase of the treclaim, so any memory written between the tbegin and the
signal will be rolled back anyway.

For signals taken in non-TM or suspended mode, we use the
normal/non-checkpointed stack pointer.

Tested with 64 and 32 bit signals

Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Cc: &lt;stable@vger.kernel.org&gt; # v3.9
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
When in an active transaction that takes a signal, we need to be careful with
the stack.  It's possible that the stack has moved back up after the tbegin.
The obvious case here is when the tbegin is called inside a function that
returns before a tend.  In this case, the stack is part of the checkpointed
transactional memory state.  If we write over this non transactionally or in
suspend, we are in trouble because if we get a tm abort, the program counter
and stack pointer will be back at the tbegin but our in memory stack won't be
valid anymore.

To avoid this, when taking a signal in an active transaction, we need to use
the stack pointer from the checkpointed state, rather than the speculated
state.  This ensures that the signal context (written tm suspended) will be
written below the stack required for the rollback.  The transaction is aborted
becuase of the treclaim, so any memory written between the tbegin and the
signal will be rolled back anyway.

For signals taken in non-TM or suspended mode, we use the
normal/non-checkpointed stack pointer.

Tested with 64 and 32 bit signals

Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Cc: &lt;stable@vger.kernel.org&gt; # v3.9
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: Context switch more PMU related SPRs</title>
<updated>2013-05-24T08:13:45+00:00</updated>
<author>
<name>Michael Ellerman</name>
<email>michael@ellerman.id.au</email>
</author>
<published>2013-05-21T16:31:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=59affcd3e460b97492bc1aa2b843bafe7c54f596'/>
<id>59affcd3e460b97492bc1aa2b843bafe7c54f596</id>
<content type='text'>
In commit 9353374 "Context switch the new EBB SPRs" we added support for
context switching some new EBB SPRs. However despite four of us signing
off on that patch we missed some. To be fair these are not actually new
SPRs, but they are now potentially user accessible so need to be context
switched.

Signed-off-by: Michael Ellerman &lt;michael@ellerman.id.au&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
In commit 9353374 "Context switch the new EBB SPRs" we added support for
context switching some new EBB SPRs. However despite four of us signing
off on that patch we missed some. To be fair these are not actually new
SPRs, but they are now potentially user accessible so need to be context
switched.

Signed-off-by: Michael Ellerman &lt;michael@ellerman.id.au&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: Context switch the new EBB SPRs</title>
<updated>2013-05-02T00:37:36+00:00</updated>
<author>
<name>Michael Ellerman</name>
<email>michael@ellerman.id.au</email>
</author>
<published>2013-04-30T20:17:04+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=9353374b8e1585d5fa47a1e5c1d3e9155dd0eb7c'/>
<id>9353374b8e1585d5fa47a1e5c1d3e9155dd0eb7c</id>
<content type='text'>
This context switches the new Event Based Branching (EBB) SPRs.  The three new
SPRs are:
  - Event Based Branch Handler Register (EBBHR)
  - Event Based Branch Return Register (EBBRR)
  - Branch Event Status and Control Register (BESCR)

Signed-off-by: Michael Ellerman &lt;michael@ellerman.id.au&gt;
Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This context switches the new Event Based Branching (EBB) SPRs.  The three new
SPRs are:
  - Event Based Branch Handler Register (EBBHR)
  - Event Based Branch Return Register (EBBRR)
  - Branch Event Status and Control Register (BESCR)

Signed-off-by: Michael Ellerman &lt;michael@ellerman.id.au&gt;
Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: remove PReP platform</title>
<updated>2013-04-18T03:03:53+00:00</updated>
<author>
<name>Paul Bolle</name>
<email>pebolle@tiscali.nl</email>
</author>
<published>2013-03-27T00:47:03+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=933ee7119fb14156f46dc8bce8218f62db13c568'/>
<id>933ee7119fb14156f46dc8bce8218f62db13c568</id>
<content type='text'>
PPC_PREP is marked as BROKEN since v2.6.15. Remove all PReP specific
code now.

Signed-off-by: Paul Bolle &lt;pebolle@tiscali.nl&gt;
Signed-off-by: Michael Ellerman &lt;michael@ellerman.id.au&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
PPC_PREP is marked as BROKEN since v2.6.15. Remove all PReP specific
code now.

Signed-off-by: Paul Bolle &lt;pebolle@tiscali.nl&gt;
Signed-off-by: Michael Ellerman &lt;michael@ellerman.id.au&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: New macros for transactional memory support</title>
<updated>2013-02-15T05:58:50+00:00</updated>
<author>
<name>Michael Neuling</name>
<email>mikey@neuling.org</email>
</author>
<published>2013-02-13T16:21:32+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=8b3c34cf0e0ab334a24aad7367cd06a5ba09a898'/>
<id>8b3c34cf0e0ab334a24aad7367cd06a5ba09a898</id>
<content type='text'>
This adds new macros for saving and restoring checkpointed architected state
from and to the thread_struct.

It also adds some debugging macros for when your brain explodes trying to debug
your transactional memory enabled kernel.

Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This adds new macros for saving and restoring checkpointed architected state
from and to the thread_struct.

It also adds some debugging macros for when your brain explodes trying to debug
your transactional memory enabled kernel.

Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: Add additional state needed for transactional memory to thread struct</title>
<updated>2013-02-15T05:58:50+00:00</updated>
<author>
<name>Michael Neuling</name>
<email>mikey@neuling.org</email>
</author>
<published>2013-02-13T16:21:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=f4c3aff2230bfe696a0683073b77446248d7895c'/>
<id>f4c3aff2230bfe696a0683073b77446248d7895c</id>
<content type='text'>
Set of new archtected state for saving away on context switch.

Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Set of new archtected state for saving away on context switch.

Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: Add support for context switching the TAR register</title>
<updated>2013-02-08T03:05:50+00:00</updated>
<author>
<name>Ian Munsie</name>
<email>imunsie@au1.ibm.com</email>
</author>
<published>2013-02-07T15:46:58+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=2468dcf641e4f3e1b0153e3e11ca20740b2f4ce8'/>
<id>2468dcf641e4f3e1b0153e3e11ca20740b2f4ce8</id>
<content type='text'>
This patch adds support for enabling and context switching the Target
Address Register in Power8. The TAR is a new special purpose register
that can be used for computed branches with the bctar[l] (branch
conditional to TAR) instruction in the same manner as the count and link
registers.

Signed-off-by: Ian Munsie &lt;imunsie@au1.ibm.com&gt;
Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This patch adds support for enabling and context switching the Target
Address Register in Power8. The TAR is a new special purpose register
that can be used for computed branches with the bctar[l] (branch
conditional to TAR) instruction in the same manner as the count and link
registers.

Signed-off-by: Ian Munsie &lt;imunsie@au1.ibm.com&gt;
Signed-off-by: Matt Evans &lt;matt@ozlabs.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: Hardware breakpoints rewrite to handle non DABR breakpoint registers</title>
<updated>2013-01-10T06:01:44+00:00</updated>
<author>
<name>Michael Neuling</name>
<email>mikey@neuling.org</email>
</author>
<published>2012-12-20T14:06:44+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=9422de3e953d0e60eb95f5430a9dd803eec1c6d7'/>
<id>9422de3e953d0e60eb95f5430a9dd803eec1c6d7</id>
<content type='text'>
This is a rewrite so that we don't assume we are using the DABR throughout the
code.  We now use the arch_hw_breakpoint to store the breakpoint in a generic
manner in the thread_struct, rather than storing the raw DABR value.

The ptrace GET/SET_DEBUGREG interface currently passes the raw DABR in from
userspace.  We keep this functionality, so that future changes (like the POWER8
DAWR), will still fake the DABR to userspace.

Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
This is a rewrite so that we don't assume we are using the DABR throughout the
code.  We now use the arch_hw_breakpoint to store the breakpoint in a generic
manner in the thread_struct, rather than storing the raw DABR value.

The ptrace GET/SET_DEBUGREG interface currently passes the raw DABR in from
userspace.  We keep this functionality, so that future changes (like the POWER8
DAWR), will still fake the DABR to userspace.

Signed-off-by: Michael Neuling &lt;mikey@neuling.org&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>powerpc: Define ppr in thread_struct</title>
<updated>2013-01-10T06:01:08+00:00</updated>
<author>
<name>Haren Myneni</name>
<email>haren@linux.vnet.ibm.com</email>
</author>
<published>2012-12-06T21:49:56+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=92779245599bb3d7fb48066b11c4bfd6aa477198'/>
<id>92779245599bb3d7fb48066b11c4bfd6aa477198</id>
<content type='text'>
[PATCH 4/6] powerpc: Define ppr in thread_struct

ppr in thread_struct is used to save PPR and restore it before process exits
from kernel.

This patch sets the default priority to 3 when tasks are created such
that users can use 4 for higher priority tasks.

Signed-off-by: Haren Myneni &lt;haren@us.ibm.com&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
[PATCH 4/6] powerpc: Define ppr in thread_struct

ppr in thread_struct is used to save PPR and restore it before process exits
from kernel.

This patch sets the default priority to 3 when tasks are created such
that users can use 4 for higher priority tasks.

Signed-off-by: Haren Myneni &lt;haren@us.ibm.com&gt;
Signed-off-by: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
</pre>
</div>
</content>
</entry>
</feed>
